Pre Work

Load Libraries and Define Custom Functions

1. Load and Inspect Data

2.A EDA - Univariate Analysis

2.A (i) Credit Score

2.A (ii) Age

2.A (iii) Estimated Salary

2.A (iv) Balance

2.A (v) Tenure

2.A (vi) NumOfProducts

2.A (vii) HasCrCard

2.A (viii) IsActiveMember

2.A (ix) Exited (Target Variable)

2.A (x) Geography

2.A (xi) Gender

2.B Bivariate Analysis

Examining bivariate relationships - vs. Exited

Bivariate vs. Balance

3. Data Preprocessing

4. Modelling

Function and Constants needed for section

4.A Base model - only one hidden layer and few nodes

4.B Refining the ANN Model - using class weights and more epochs

4.C Using Tanh as activation for the hidden layers

4.D Trying more complex model -

Final Conclusions and Business Recommendations

Appendix: Using XGBoost to check ANN performance